摘要
空中目标的逆合成孔经雷达(inverse synthetic aperture radar,ISAR)成像效果往往不理想,干扰杂波、成像算法能力限制等因素使得目标图像质量较差,较难从图像中提取目标的连续轮廓信息。对此,提出一种新的基于轮廓波降噪的处理方法以获得目标连续轮廓特征。首先应用轮廓波变换将复数ISAR图像变换为轮廓波系数,在轮廓波变换域分离信号与噪声并完成降噪处理,接着用优化后的轮廓波系数重建ISAR复图像,然后在实图像域将用形态学方法获得目标的初始轮廓改进为CV(Chan-Vese)模型算法的初始轮廓,最后用CV模型算法进行有限次迭代以获得目标的较优连续轮廓特征。该方法融合了轮廓波降噪技术与自适应CV模型算法,通过对真实yak42型飞机目标ISAR回波数据的试验验证了该方法的有效性和可行性。
Since interference noise and capability limit of imaging algorithm and other factors degrade image quality, it is difficult to extract target continuous contour information from the image; therefore the inverse synthetic aperture radar （ISAR） imaging effect of aerial targets is not ideal. Aiming at this problem, a new image processing method based on contourlet noise reduction is proposed to achieve clearer continuous contour features of the target. Firstly, complex ISAR image is transformed into contourlet coefficients by means of contourlet transform. The contourlet coefficients are denoised and complex ISAR image is reconstructed from the denoised contourlet coefficients. Initial contour of the target is obtained with morphology method and then Chan-Vese （CV） model method is applied to the ini- tial contour, and better continuous contour features of the target are obtained with finite iterations. The proposed method combines the contourlet noise reduction technology and adaptive CV model algorithm. Real yak42 aircraft was tested; and the ISAR echo data of the target verify the validity and feasibility of this method.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2012年第10期2293-2300,共8页
Chinese Journal of Scientific Instrument
基金
军口863项目(2011SQ7072054)
教育部博士点基金(1020607220090142)
中国工程物理研究院院重点基金(2010A0403d7)资助项目
关键词
逆合成孔径雷达
特征提取
轮廓波
形态学
CV算法
inverse synthetic aperture radar
feature extraction
contourlet
morphology
Chan-Vese（CV） algorithm
